Journal of the Civil Engineering Forum

Effect of Relative Density Variation on The Effectiveness of Fly Ash in Stabilization of Sandy Soils for Liquefaction Mitigation

Yunan Yakuta Wangsawitana — 2026-05-04

Indonesia is highly vulnerable to earthquakes that can trigger soil liquefaction, particularly in coastal regions where saturated sandy soils with low relative densities are prevalent. This study investigated the influence of varying Class C fly ash contents (5%, 10%, and 15%), relative densities (50%, 70%, and 90%), and curing periods (7 and 14 days) on the unconfined compressive strength of fully saturated sandy soils as a liquefaction mitigation approach. The methodology included physical soil characterization, sample preparation with specific fly ash and relative density variations, and unconfined compressive strength testing after the curing period. The results demonstrated that increasing the fly ash content, relative density, and curing time significantly enhanced the strength of the samples. The highest strength occurred at 15% fly ash and 90% relative density, with 426.22 kPa (7 days) and 438.57 kPa (14 days). The most notable improvement occurred with an increase in fly ash content from 5% to 10% and relative density from 50% to 70%. Fly ash was more effective at enhancing the strength of sandy soil by promoting interparticle cementation, whereas excessive compaction tended to disrupt the soil structure and generate fine particles that reduced cohesion, as evidenced by the UCS test results and SEM observations. The most significant strength gain between 7 and 14 days was observed at 50% relative density and 5% fly ash, with a twofold increase. The combination of 10% fly ash and 70% relative density was optimal for field implementation, as it yielded strength values exceeding the liquefaction resistance thresholds, while maintaining practical compaction levels and material efficiency. A comparative cost assessment considering materials, labor, equipment, and construction time is recommended to evaluate the practical feasibility of fly ash stabilization relative to alternative ground improvement methods for Indonesian projects.

Real-time Application and Statistical Evaluation of Reservoir Operating Rules Under Inflow Uncertainty in Monte Carlo Simulations

Muhammad Mujiburrakhman — 2026-06-18

Reservoir Operating Rules (ROR) have been well-known in Indonesia as guidelines for reservoir operation that comprise the release rules and rule curves derived from the dependable inflow time series in dry, normal, and wet conditions. However, likely because of the incoherence of the original application concept of ROR, the real performance of ROR in reservoir operation has so far remained unclear. This paper proposes a practical application concept of ROR and evaluates the performance of its implementation for annual storage conservation and periodic demand satisfaction under inflow uncertainty in the integrated system of Segawe Regulating Weir, Wonorejo Multipurpose Dam, and Tiudan Regulating Dam. The release rules for the three hydraulic structures were integrated into a unified “Wonorejo ROR”throughoptimization of water allocation over 36 ten-day operational periods. The performance of the application of Wonorejo ROR was evaluated based on several operational success and failure events in Monte Carlo simulations, for which 1,000 years of unique and independent synthetic inflow time series were generated from the Thomas–Fiering model. Under the application concept of ROR, the release decisions for the three hydraulic structures were made in real-time solely based on the actual water level of Wonorejo Reservoir and the rule curves of Wonorejo ROR. For comparative purposes, the Monte Carlo simulations were carried out to evaluate two alternatives of Wonorejo ROR, namely ROR-D65 and ROR-D80, which had different dry inflow exceedance probabilities. Based on the specified events in the Monte Carlo simulations, ROR-D65 and ROR-D80 resulted in reliabilities of 76.4% and 81.3% for annual storage conservation, and the lowest reliabilities of 73.9% and 52.4% for periodic demand satisfaction, respectively. In addition, other reservoir operational performancesandacomprehensive interpretation oftheproposed applicationconcept of ROR arepresented in this paper.

Analyzing Trends and Framework Models for Autonomous Vehicle Transition Strategies

Cintia Nurliyana — 2026-07-01

The shift from traditional human-driven to autonomous vehicles represents a transformation that is not merely technological, but also social. Currently, the deployment of autonomous vehicles remains largely at level where the human driver continues to be the primary decisionmaker. However, this dynamic is expected to change significantly as higher levels of autonomy are achieved. This study aims to identify the impacts that occur during process transition through emerging trends and theoretical lens of transition studies. The initial stage involved mapping trends in Autonomous Vehicle (AV) research to identify key related topics. This study uses bibliometric analysis of 82 key articles selected through a screening process in a systematic review. VOSviewer and CiteSpace were employed to map the research landscape, identify dominant topics, and trace the evolution of AV studies over time. The analysis highlights both well-established research areas and underexplored gaps. Furthermore, the articles were classified into categories based on transition study dimensions and key actors involedinvolved in facilitating a successful transition. The findings reveal a dominant research focus on users, human factors, and mobility behavior, whereas studies concerning vulnerable road users and the industry transition remain limited. To address this gap, this study investigates the socio-technical factors arising from the identified trends and transition studies while also exploring the roles of key stakeholders. This approach aims to achieve a more balanced perspective. Specifically, the study proposes a novel framework that connects critical actors with core socio-technical dimensions to support a sustainable autonomous ve-hicle transition. This research thus provides a strategic framework designed to guide and accelerate the shift toward autonomous mobility.

Modeling Acidity Dynamics Driven by Tidal–Rainfall Interactions in a Tropical Lowland Irrigation Network

Samuel Harjanto — 2026-07-17

Water acidity is a critical determinant of rice productivity in tidal swamp irrigation systems. In the Terusan Tengah Irrigation Area, irrigation water is characterized by high acidity, conditions that may adversely impact rice cultivation. Although extensive field measurements have confirmed the presence of acid water within the network, observational data alone cannot disentangle how acidity patterns in tidal systems are controlled by tidal advection or generated by rainfall-driven runoff. Therefore, two contrasting hydrological scenarios rain-free and post-rainfall were simulated to isolate the mechanism governing pH dynamics. This study employs conservative hydrodynamic modelling using HEC-RAS, representing pH as hydrogen ion concentration, to investigate shifts in the controlling mechanisms of acidity under two contrasting hydrological conditions: rain-free and rainfall-affected periods. The simulations were designed to distinguish pH responses driven by tidal transport from those triggered by rainfall. Under rain-free conditions, tidal inflow transports acidic water into the primary canal, causing acidity to accumulate near segments underlain by shallow pyrite layers, but pH partially recovers during ebb tide as the acidic mass is flushed seaward. In contrast, after rainfall, runoff-driven acidity converges toward the estuary and persists even during low tide, as elevated water levels and pseudo-tidal retention inhibit flushing and prolong the residence time of acidic water within the network. This behavior reflects a governing-mechanism transition: tides act merely as distributors of existing acidity during dry periods, whereas rainfall initiates acid generation and retention within the network, concentrated near the estuary. These findings enhance the understanding of acidity dynamics in the Terusan Tengah irrigation network—driven by tidal fluctuations and rainfall—and provide an initial basis for developing adaptive operational strategies to mitigate acidity accumulation and its impacts on ecosystems, crop productivity, and local livelihoods.

Speed and Lane Change Characteristics of Motorcycles at Priority Junction

Benidiktus Susanto — 2026-07-20

Motorcycles are Indonesia’s predominant mode of transportation and contribute to a substantial share of road traffic accidents. The research aims to identify key behavioural factors contributing to motorcycle-related conflict risks at intersections and to propose evidence-based safety interventions. This study analyzed speed characteristics with varying degrees of disruption that motorists would experience at the intersection. These disruptions were categorized into high, low, no disruption, and left-turn movements. High disruption was defined as vehicles stopping in the middle of the intersection as motorcyclists approached, while low disruption was defined as vehicles preparing to enter the intersection from minor roads. In addition to speed characteristics, this study also observed lane-changing behavior when disruptions occurred at the intersection. To obtain detailed data, drone-based cameras were used for observations, and the results were validated against radar speed measurements. The results show that 67.94% of riders exceed the 60 km/h speed limit during the morning period, while 33.27% exceed it in the afternoon, indicating clear temporal differences in operating speeds. Approximately 85 per cent of lane-changing manoeuvres occur within 20 metres of the conflict point, reflecting short decision distances that increase collision risk. Riders also exhibit assertive gap-acceptance behaviour, often requiring main-road vehicles to adjust their speed. Statistical validation confirms that there is no significant difference between drone-derived and radar-based speed measurements, demonstrating the reliability of the drone-based methodology. Based on these findings, the study recommends targeted speed management measures, including automated enforcement and the use of traffic-calming devices; improved lane discipline through more precise road markings and selective physical separation; and junction design enhancements, such as motorcycle waiting areas and advanced stop lines. These measures can improve motorcycle safety at priority junctions and support more efficient and predictable urban traffic operations.

How Toll Roads Reshape Structural Vulnerability in Primary Road Networks: A Graph Neural Network Learning from Central Java Province

Berlian Kushari — 2026-07-20

The roles of primary road networks in maintaining regional connectivity, supporting economic activities, and enabling emergency response, especially in disaster-prone areas, are well known. Vulnerability analysis of such networks is commonly conducted following disruption-based scenarios, in which external hazards are modeled to disrupt and degrade network performance. This study adopts a different perspective by considering vulnerability as a property that can also arise from the network’s own topological structure. In this approach, vulnerability is assumed to exist even prior to external disturbances, and thus is an intrinsic characteristic of the network. To explore this idea, a Graph Neural Network (GNN)-based framework is proposed to learn structural vulnerability directly from network topology. The primary road network of Central Java Province, Indonesia, was used as a case study. The network was modeled as an undirected, weighted graph representing two configurations: without toll roads and with toll road integration. Vulnerability labels at the node level were derived from a series of node removal experiments, capturing efficiency loss, connectivity degradation, and network fragmentation. A Graph Convolutional Network (GCN) model with two GCN layers was then trained to learn patterns associated with structural vulnerability based on node features and graph connectivity. The results indicate that structural vulnerability can be learned as a network property, with the model achieving an accuracy of approximately 82% across both configurations. This performance exceeds that of conventional centrality-based approximations. The findings also show that the inclusion of toll roads reshapes the distribution of vulnerability across the network. This supports the interpretation that toll roads act as structural modifiers by reconfiguring systemic vulnerability rather than simply increasing redundancy.

Daily Water Balance Analysis of the Sempor Reservoir in Central Java Using the SWAT Model for Operational Reliability Assessment

Ayu Septyandini Yudiva Putri — 2026-07-30

empor Reservoir plays a crucial role in supporting irrigation water supply, raw water provision, and flood control in Central Java. Effective reservoir operation requires reliable estimates of inflow and water availability to ensure sustainable water allocation under varying hydrological conditions. This study analyzes the daily water balance of Sempor Reservoir using the Soil and Water Assessment Tool (SWAT) model to simulate inflow and assess operational reliability. Daily rainfall and climate data for 2015–2024 were used, with streamflow observations used for model calibration (2015–2022) and validation (2023–2024). Model performance was evaluated using the Nash–Sutcliffe Efficiency (NSE) and the coefficient of determination (R²). The SWAT model performed very well, achieving NSE = 0.83 and R2 = 0.87. These performance statistics indicate that the simulated inflows closely matched the observed discharge and provide confidence in the model’s application for water balance analysis. Daily water balance results show that deficit conditions occur mainly during the dry season. Annual inflow and demand were 184.02 million m³ and 186.45 million m³, resulting in a deficit of 2.43 million m³. The reservoir was full for 36 days, sufficient for 311 days, and insufficient for 18 days, indicating approximately 95% reliability. Overall, the results demonstrate that the SWAT model is an effective tool for evaluating daily reservoir water balance and operational performance. The findings provide useful information for reservoir operation, water allocation planning, and sustainable water resources management under changing hydrological conditions. The results also provide valuable information for water managers in optimizing reservoir operation, improving water allocation efficiency, and reducing the impacts of seasonal water shortages.

Scenario-Neutral Assessment of Rainfall–Erosion Sensitivity Using RUSLE in the Sempor Reservoir Catchment, Indonesia

Irfan Fadillah Widianto — 2026-07-30

Rainfall variability poses a substantial threat to reservoir sustainability in tropical monsoon catchments because changes in rainfall intensity and seasonality directly influence soil detachment, runoff generation, and sediment source pressure. Conventional climate–impact assessments commonly rely on a limited number of deterministic climate projections, which may obscure the broader sensitivity of catchment erosion to plausible rainfall changes. This study integrates a scenario–neutral (SN) framework with the Revised Universal Soil Loss Equation (RUSLE) to evaluate rainfall–erosion sensitivity in the Sempor Reservoir catchment, Central Java, Indonesia. Historical daily rainfall was used to establish the baseline, while systematic perturbations of annual precipitation, seasonal rainfall distribution, wet–day frequency, and extreme rainfall intensity were used to construct a broad rainfall–exposure space. Changes in rainfall erosivity were subsequently propagated through RUSLE while soil erodibility, topography, land cover, and support–practice factors were held constant. The resulting estimates represent potential hillslope soil loss and sediment–source pressure rather than direct measurements of sediment delivery or reservoir deposition. The results reveal pronounced seasonal and spatial contrasts. Wet–season rainfall contributes more than 60% of annual rainfall erosivity and approximately 70–80% of annual estimated soil loss. The greatest estimated erosion occurs where relatively high rainfall erosivity coincides with higher soil–erodibility, topographic, cover–management, and support practice factors. Increasing wet–season rainfall produces a marked increase in estimated soil loss, indicating that erosion sensitivity is influenced by rainfall timing and intensity as well as annual rainfall totals. The scenario neutral simulations reproduce baseline medians and seasonal patterns consistently but attenuate the upper distribution tails and consequently underrepresent rare extreme events. The integrated SN–RUSLE framework therefore provides a transparent approach for diagnosing rainfall–erosion sensitivity under climate uncertainty. The findings support prioritising wet–season erosion control, maintaining vegetation cover, and implementing spatially targeted conservation measures in areas with higher estimated erosion potential.

Weathering-Induced Strength Degradation and Hydro-Mechanical Controls on the Stability of Multilayered Tropical Rock Slopes

Stephanie Ivy Luna — 2026-08-05

Tropical environments accelerate rock weathering, which can significantly reduce the stability of mineralized slopes. This study investigates the influence of weathering and groundwater conditions on the stability of a strong–weak–strong multilayered rock slope hosting an iron deposit in the Philippines, where intense rainfall and highly fractured rock masses are common. Detailed geological and hydrogeological investigations were conducted to construct two-dimensional numerical models of the slope system. Mechanical characterization of both weathered and unweathered rock masses was performed using field hardness testing and laboratory
geomechanical tests. These parameters were incorporated into numerical simulations to evaluate slope stability under varying degrees of weathering, weak-layer thickness, and hydrological conditions derived from developed phreatic surfaces. Results indicate that when the middle weak layer is thin, a 50% weathering intensity still allows a stable multilayered slope at an Overall Slope Angle (OSA) of 60°. However, as weak-layer thickness increases, the same weathering degree leads to slope instability. Weathering effects recorded at the
investigated study area correspond to an average reduction of 10.8 units in the Geological Strength Index (GSI), and are equivalent to a 50% weathering degree, implying that each 0.2 reduction in GSI represents approximately 1% weathering of the rock mass. Considering weathering-induced strength degradation, the optimally safe slope angle decreases from the originally designed 60° to approximately 35°, maintaining a minimum factor of safety of 1.2. Although the stabilizing effect of other countermeasures was excluded from this study. Hydrological analysis further reveals that low rainfall may temporarily increase slope stability due to matric suction induced by capillary effects in clay-like mineral sericite, although prolonged rainfall generates positive pore pressures that reduce stability. The thickness of the weak layer also produces localized increases in hydraulic gradient due to permeability contrasts, increasing susceptibility to rainfall-induced failure. These findings highlight the importance of quantitatively incorporating weathering and hydrogeological effects in slope design for tropical open-pit mines.

Evaluating Slope Instability in a Volcanic Caldera Environment Using GIS-Based 3D Analysis: Mt. Bawakaraeng, South Sulawesi

Poppy Indrayani — 2026-08-24

Landslides in volcanic terrains pose significant hazards to downstream communities, infrastructure, and watershed systems, particularly where steep slopes and complex geological conditions coexist. This study presents a GIS-based three-dimensional (3D) deterministic slope stability analysis to evaluate landslide susceptibility within the Mt. Bawakaraeng caldera, Sulawesi Island, Indonesia. The analysis integrates geotechnical, geological, hydrological, and topographic parameters obtained from field investigations, laboratory testing, remote sensing data, and a high-resolution 5-m Digital Elevation Model (DEM). Slope stability was evaluated using the factor of safety (FS), calculated through a Hovland-based 3D column method implemented within a GIS environment. Two critical collapse zones were identified and corresponded closely with documented historical landslide events. The calculated minimum FS values of 0.825 and 1.000 indicate unstable to marginally stable slope conditions, respectively. Numerical simulations indicate a progressive failure mechanism in which the collapse of a major slope unit can reduce the stability of adjacent slope units, potentially generating cascading landslide events. Monte Carlo simulations were additionally employed to identify critical slip surfaces and assess potential failure configurations. The simulated slip surfaces showed good agreement with observed landslide morphologies, supporting the reliability of the modeling approach. Potential failure volumes were also estimated to assess the downstream sediment hazard to the Jeneberang watershed. The results demonstrate that integrating GIS, high-resolution terrain data, and 3D deterministic modeling provides an effective framework for spatially explicit slope stability assessment in complex volcanic environments. The approach can support landslide hazard mitigation, watershed management, infrastructure planning, and sustainable development in mountainous regions prone to slope failure.